Towards trustworthy multi-modal motion prediction:Holistic evaluation and interpretability of outputs

Predicting the motion of other road agents enables autonomous vehicles to perform safe and efficient path planning.This task is very complex,as the behaviour of road agents depends on many factors and the number of possible future trajectories can be consid-erable(multi-modal).Most prior approaches proposed to address multi-modal motion prediction are based on complex machine learning systems that have limited interpret-ability.Moreover,the metrics used in current benchmarks do not evaluate all aspects of the problem,such as the diversity and admissibility of the output.The authors aim to advance towards the design of trustworthy motion prediction systems,based on some of the re-quirements for the design of Trustworthy Artificial Intelligence.The focus is on evaluation criteria,robustness,and interpretability of outputs.First,the evaluation metrics are comprehensively analysed,the main gaps of current benchmarks are identified,and a new holistic evaluation framework is proposed.Then,a method for the assessment of spatial and temporal robustness is introduced by simulating noise in the perception system.To enhance the interpretability of the outputs and generate more balanced results in the proposed evaluation framework,an intent prediction layer that can be attached to multi-modal motion prediction models is proposed.The effectiveness of this approach is assessed through a survey that explores different elements in the visualisation of the multi-modal trajectories and intentions.The proposed approach and findings make a significant contribution to the development of trustworthy motion prediction systems for autono-mous vehicles,advancing the field towards greater safety and reliability.

A review of electrospun separators for lithium-based batteries: Progress and application prospects

Due to the limitations of the raw materials and processes involved,polyolefin separators used in commercial lithium-ion batteries(LIBs)have gradually failed to meet the increasing requirements of high-end batteries in terms of energy density,power density,and safety.Hence,it is very important to develop next-generation separators for advanced lithium(Li)-based recharge-able batteries including LIBs and Li-S batteries.Nonwoven nanofiber membranes fabricated via electrospinning technology are highly attractive candidates for high-end separators due to their simple processes,low-cost equipment,controllable microporous structure,wide material applicability,and availability of multiple functions.In this review,the electrospinning technologies for separators are reviewed in terms of devices,process and environment,and polymer solution systems.Furthermore,strategies toward the improvement of electrospun separators in advanced LIBs and Li-S batteries are presented in terms of the compositions and the structure of nanofibers and separators.Finally,the challenges and prospects of electrospun separators in both academia and industry are proposed.We anticipate that these systematic discussions can provide information in terms of commercial applications of electrospun separators and offer new perspectives for the design of functional electrospun separators for advanced Li-based batteries.

Obituary:Prof.Yun Zhang(1963-2023)-A scientist focused on toxins and their underlying mechanisms to decipher human diseases

Prof.Yun Zhang was born on 9 July 1963 in Kunming,Yunnan,China,during a tumultuous period which he often referenced.Throughout his life,he harbored a steadfast belief in using knowledge to unravel the mysteries of human diseases.His educational journey was marked by frequent changes in schools due to his parents’occupational relocations.However,despite these challenges,he consistently displayed diligence and was admitted to the East China University of Science and Technology,Shanghai,after completing high school in 1980.He remained an active and loyal member of the School of Biotechnology at the university.

Impact of oxygen incorporation on interface optimization and defect suppression for efficiency enhancement in Cu_(2)ZnSn(S,Se)_(4)solar cells

Kesterite Cu_(2)ZnSn(S,Se)_(4)(CZTSSe)solar cells suffer from severe carrier recombination,limiting the photovoltaic performance.Unfavorable energy band alignment at the p-n junction and defective front interface are two main causes.Herein,oxygen incorporation in CZTSSe via absorber air-annealing was developed as a strategy to optimize its surface photoelectric property and reduce the defects.With optimized oxygen incorporation conditions,the carrier separation and collection behavior at the front interface of the device is improved.In particular,it is found that oxygen incorporated absorber exhibits increased band bending,larger depletion region width,and suppressed absorber defects.These indicate the dynamic factors for carrier separation become stronger.Meanwhile,the increased potential difference between grain boundaries and intra grains combined with the decreased concentration of interface deep level defect in the absorber provide a better path for carrier transport.As a consequence,the champion efficiency of CZTSSe solar cells has been improved from 9.74%to 12.04%with significantly improved open-circuit voltage after optimized air-annealing condition.This work provides a new insight for interface engineering to improve the photoelectric conversion efficiency of CZTSSe devices.

Unraveling the efficiency losses and improving methods in quantum dot-based infrared up-conversion photodetectors

Quantum dot-based up-conversion photodetector,in which an infrared photodiode(PD)and a quantum dot light-emitting diode(QLED)are back-to-back connected,is a promising candidate for low-cost infrared imaging.However,the huge efficiency losses caused by integrating the PD and QLED together hasn’t been studied sufficiently.This work revealed at least three origins for the efficiency losses.First,the PD unit and QLED unit usually didn’t work under optimal conditions at the same time.Second,the potential barriers and traps at the interconnection between PD and QLED units induced unfavorable carrier recombination.Third,much emitted visible light was lost due to the strong visible absorption in the PD unit.Based on the understandings on the loss mechanisms,the infrared up-conversion photodetectors were optimized and achieved a breakthrough photon-to-photon conversion efficiency of 6.9%.This study provided valuable guidance on how to optimize the way of integration for up-conversion photodetectors.

列车通过分岔隧道的气动效应研究

长大铁路隧道受特殊地质条件影响可采用分岔隧道建造。然而,列车高速穿越这种复杂构型的隧道时所引起的空气动力学效应尚不清楚。为此,本文采用三维、可压缩、非定常的雷诺时均模型,模拟研究了高速列车从不同方向通过分岔隧道时引起的气动效应。研究内容主要包括入口隧道结构为单线隧道(STE)和双线隧道(DTE)两个运行环境下的车体表面和隧道壁面的压力波动、列车气动力和周围流场特征等。研究结果表明,相比STE场景,DTE场景下车体表面和单线隧道壁面的压力峰峰值更大,并且气动阻力能耗也更大,头车的平均阻力增大19.2%。然而,STE场景下列车的侧向力波动更剧烈,其峰值相比DTE场景增加26%。本文的研究发现可为特殊铁路隧道的设计和建造提供有价值的空气动力学参考。

Analysis for Effects of Temperature Rise of PV Modules upon Driving Distance of Vehicle Integrated Photovoltaic Electric Vehicles

The development of vehicle integrated photovoltaics-powered electric vehicles (VIPV-EV) significantly reduces CO2 emissions from the transport sector to realize a decarbonized society. Although long-distance driving of VIPV-EV without electricity charging is expected in sunny regions, driving distance of VIPV-EV is affected by climate conditions such as solar irradiation and temperature rise of PV modules. In this paper, detailed analytical results for effects of climate conditions such as solar irradiation and temperature rise of PV modules upon driving distance of the VIPV-EV were presented by using test data for Toyota Prius and Nissan Van demonstration cars installed with high-efficiency InGaP/GaAs/InGaAs 3-junction solar cell modules with a module efficiency of more than 30%. The temperature rise of some PV modules studied in this study was shown to be expressed by some coefficients related to solar irradiation, wind speed and radiative cooling. The potential of VIPV-EV to be deployed in 10 major cities was also analyzed. Although sunshine cities such as Phoenix show the high reduction ratio of driving range with 17% due to temperature rise of VIPV modules, populous cities such as Tokyo show low reduction ratio of 9%. It was also shown in this paper that the difference between the driving distance of VIPV-EV driving in the morning and the afternoon is due to PV modules’ radiative cooling. In addition, the importance of heat dissipation of PV modules and the development of high-efficiency PV modules with better temperature coefficients was suggested in order to expand driving range of VIPV-EV. The effects of air-conditioner usage and partial shading in addition to the effects of temperature rise of VIPV modules were suggested as the other power losses of VIPV-EV.

Deep learning enhanced NIR-Ⅱ volumetric imaging of whole mice vasculature

Fluorescence imaging through the second near-infrared window(NIR-Ⅱ,1000–1700 nm) allows in-depth imaging.However, current imaging systems use wide-field illumination and can only provide low-contrast 2D information, without depth resolution. Here, we systematically apply a light-sheet illumination, a time-gated detection, and a deep-learning algorithm to yield high-contrast high-resolution volumetric images. To achieve a large Fo V(field of view) and minimize the scattering effect, we generate a light sheet as thin as 100.5 μm with a Rayleigh length of 8 mm to yield an axial resolution of 220 μm. To further suppress the background, we time-gate to only detect long lifetime luminescence achieving a high contrast of up to 0.45 Icontrast. To enhance the resolution, we develop an algorithm based on profile protrusions detection and a deep neural network and distinguish vasculature from a low-contrast area of 0.07 Icontrast to resolve the 100μm small vessels. The system can rapidly scan a volume of view of 75 × 55 × 20 mm3and collect 750 images within 6mins. By adding a scattering-based modality to acquire the 3D surface profile of the mice skin, we reveal the whole volumetric vasculature network with clear depth resolution within more than 1 mm from the skin. High-contrast large-scale 3D animal imaging helps us expand a new dimension in NIR-Ⅱ imaging.

3D printed fiber-optic nanomechanical bioprobe

Ultrasensitive nanomechanical instruments,e.g.atomic force microscopy(AFM),can be used to perform delicate biomechanical measurements and reveal the complex mechanical environment of biological processes.However,these instruments are limited because of their size and complex feedback system.In this study,we demonstrate a miniature fiber optical nanomechanical probe(FONP)that can be used to detect the mechanical properties of single cells and in vivo tissue measurements.A FONP that can operate in air and in liquids was developed by programming a microcantilever probe on the end face of a single-mode fiber using femtosecond laser two-photon polymerization nanolithography.To realize stiffness matching of the FONP and sample,a strategy of customizing the microcantilever’s spring constant according to the sample was proposed based on structure-correlated mechanics.As a proof-of concept,three FONPs with spring constants varying from 0.421 N m^(−1)to 52.6 N m^(−1)by more than two orders of magnitude were prepared.The highest microforce sensitivity was 54.5 nmμN^(−1)and the detection limit was 2.1 nN.The Young’s modulus of heterogeneous soft materials,such as polydimethylsiloxane,muscle tissue of living mice,onion cells,and MCF-7 cells,were successfully measured,which validating the broad applicability of this method.Our strategy provides a universal protocol for directly programming fiber-optic AFMs.Moreover,this method has no special requirements for the size and shape of living biological samples,which is infeasible when using commercial AFMs.FONP has made substantial progress in realizing basic biological discoveries,which may create new biomedical applications that cannot be realized by current AFMs.

An AuNPs/Mesoporous NiO/Nickel Foam Nanocomposite as a Miniaturized Electrode for Heavy Metal Detection in Groundwater

Heavy metals,notably Pb2+and Cu^(2+),are some of the most persistent contaminants found in groundwater.Frequent monitoring of these metals,which relies on efficient,sensitive,cost-effective,and reliable methods,is a necessity.We present a nanocomposite-based miniaturized electrode for the concurrent measurement of Pb2+and Cu^(2+)by exploiting the electroanalytical technique of square wave voltammetry.We also propose a facile in situ hydrothermal calcination method to directly grow binder-free mesoporous Ni O on a three-dimensional nickel foam,which is then electrochemically seeded with gold nanoparticles(Au NPs).The meticulous design of a low-barrier Ohmic contact between mesoporous Ni O and Au NPs facilitates target-mediated nanochannel-confined electron transfer within mesoporous Ni O.As a result,the heavy metals Pb2+(0.020 mg.L^(-1)detection limit;2.0–16.0 mg.L^(-1)detection range)and Cu^(2+)(0.013 mg.L^(-1)detection limit;0.4–12.8 mg.L^(-1)detection range)can be detected simultaneously with high precision.Furthermore,other heavy metal ions and common interfering ions found in groundwater showed negligible impacts on the electrode’s performance,and the recovery rate of groundwater samples varied between 96.3%±2.1%and 109.4%±0.6%.The compactness,flexible shape,low power consumption,and ability to remotely operate our electrode pave the way for onsite detection of heavy metals in groundwater,thereby demonstrating the potential to revolutionize the field of environmental monitoring.

Encrypted optical fiber tag based on encoded fiber Bragg grating array

Optical fibers are typically used in telecommunications services for data transmission,where the use of fiber tags is essential to distinguish between the different transmission fibers or channels and thus ensure the working functionality of the communication system.Traditional physical entity marking methods for fiber labeling are bulky,easily confused,and,most importantly,the label information can be accessed easily by all potential users.This work proposes an encrypted optical fiber tag based on an encoded fiber Bragg grating(FBG)array that is fabricated using a point-by-point femtosecond laser pulse chain inscription method.Gratings with different resonant wavelengths and reflectivities are realized by adjusting the grating period and the refractive index modulations.It is demonstrated that a binary data sequence carried by a fiber tag can be inscribed into the fiber core in the form of an FBG array,and the tag data can be encrypted through appropriate design of the spatial distributions of the FBGs with various reflection wavelengths and reflectivities.The proposed fiber tag technology can be used for applications in port identification,encrypted data storage,and transmission in fiber networks.

高速列车隧道压力波数据的分段线性表示

分段线性表示(PLR)技术目前已在多个领域被广泛用于重新表征高维的时间序列数据,以实现数据维度降低、波动过滤和整体趋势提取的目的。然而,该技术目前尚未在高速列车的隧道压力波数据的相关分析处理中得到应用。因此,本研究首次引入PLR技术对典型的高维列车压力波数据序列进行重新表征,并针对高速列车的压力波数据专门设计了一种基于感知重要点(PIP)且性能表现较好的PLR算法。研究结果表明,数据点重要性的度量方法和分段误差评估方法,特别是前者,对压力波序列的PIP的识别优先级甚至最终结果都会产生影响。与欧氏距离(ED)和正交距离(OD)相比,当将垂直距离(VD)作为数据点重要性的度量方法时,PLR_PIP算法(PLR_PIP_VD)获得了更合理的高速列车压力波的PLR结果。通过累积误差、平均误差和最大误差三种分段误差评估方法之间的对比,当将累积误差作为分段误差评估方法时,PLR_PIP_VD算法得到了相对较好的高速列车压力波的PLR结果。提出的适用于高速列车压力波PLR分析的PLR_PIP算法,为高速列车的压力波数据序列的分析处理提供了一种新的方法。

通风竖井位置对高速列车通过地下车站瞬态压力的影响试验

当列车快速驶过地下车站时,会产生强烈的瞬态压力波动。本文采用移动模型试验装置研究了列车通过地下车站时列车和屏蔽门表面的瞬态压力分布特征,分析了通风竖井位置对缓解瞬态压力波动的影响。结果表明,不同屏蔽门上对应位置处测点的压力波动值存在差异。通风竖井可以影响压力幅值,不同位置的通风竖井影响特性也有很大差异;位于站台旁边的竖井对缓解列车和屏蔽门表面的瞬态压力幅值效果最好。与没有竖井的车站相比,当竖井设置在站台附近位置时,布置在列车表面上的H2测点的压力幅值减小47.1%,站台屏蔽门表面上的P1测点的压力幅值减小71.3%。

Specialty optical fibers for advanced sensing applications

Optical fiber technology has changed the world by enabling extraordinary growth in world-wide communications and sensing.The rapid development and wide deployment of optical fiber sensors are driven by their excellent sensing performance with outstanding flexibility,functionality,and versatility.Notably,the research on specialty optical fibers is playing a critical role in enabling and proliferating the optical fiber sensing applications.This paper overviews recent developments in specialty optical fibers and their sensing applications.The specialty optical fibers are reviewed based on their innovations in special structures,special materials,and technologies to realize lab in/on a fiber.An overview of sensing applications in various fields is presented.The prospects and emerging research areas of specialty optical fibers are also discussed.

Numerical study on dynamic properties of rubberised concrete with different rubber contents

As a green environmentally-friendly material,rubberised concrete(Ru C),which has the characteristics of low elastic modulus,large deformation capacity,high damping,good energy dissipation and good crack resistance,has attracted extensive attention and research in the civil engineering discipline.However,most of existing studies are based on experimental tests on Ru C material properties,and there has been no numerical study based on meso-scale modelling of Ru C yet.To more comprehensively investigate the Ru C dynamic material properties without conducting intensive experimental tests,this study developed a high-fidelity meso-scale model considering coarse and fine aggregates and rubber crumbs to numerically investigate the mechanical properties of rubberised concrete under different strain rates.The meso-scale model was verified against both quasi-static compressive testing data and Split Hopkinson Pressure Bar(SHPB)dynamic testing data.Using the verified numerical model,the dynamic properties of rubberised concrete with various rubber content(0%-30%)under different strain rates were studied.The numerical results show that the developed meso-scale model can use to predict the static and dynamic properties of rubberised concrete with high accuracy.The dynamic compressive strength of the rubberised concrete increases with the increment of the strain rate,and the strain rate sensitivity increases with the rubber content ranging from 0 to 30%.Based on intensive numerical simulation data,empirical DIFs is used as a function of strain rate and rubber content to predict the dynamic strength of rubberised concrete.

Ethylenediamine tetramethylenephosphonic acid as a selective collector for the improved separation of chalcopyrite against pyrite at low alkalinity

Chalcopyrite is the main Cu-containing mineral and cannot be separated well from pyrite using traditional xanthate collectors with large amounts of lime depressant, resulting in difficulties of the tailing treatment and associated precious metals recovery. Therefore, in this study, the green and odourless ethylenediamine tetramethylenephosphonic acid(EDTMPA) was introduced as a novel chalcopyrite collector. Flotation results from the binary mineral mixture and real ore demonstrated that EDTMPA could realize the selective separation of chalcopyrite from pyrite relative to ethyl xanthate(EX) without any depressants within the wide p H range of 6.0–11.0, and might replace the traditional high-alkaline lime process. Electrochemical and Fourier transform infrared spectra measurements indicated that the difference in adsorption performance of EDTMPA on chalcopyrite and pyrite was larger than that of EX, suggesting a better selectivity for EDTMPA. Density functional theory calculations demonstrated that there were stronger chemical bonds between P—O groups of EDTMPA and the Fe/Cu atoms on chalcopyrite in the form of a stable six-membered ring. Crystal chemistry calculations further revealed that the activity of metal atoms of chalcopyrite was higher than that of pyrite. Therefore, these basic theoretical results and practical application provide a guidance for the industrial application of EDTMPA in chalcopyrite flotation.

Rigid block characteristics on subaerial landslide-tsunamis using a 3D coupled Eulerian-Lagrangian model

To quantitatively reveal how rock blocks falling into water affect the impulse waves,the influence of a rigid block on induced first wave and second wave is systematically investigated.The block characteristics include the initial velocity,density,volume,and incident angle,and the investigated wave behavior characteristics include the maximum kinetic energy of the water,the transformation ratio of the kinetic energy from the block to the waves,the duration of the waves,the maximum movement speed,and the maximum height and width of the waves.The coupled Eulerian-Lagrangian method(CEL)is introduced to establish the numerical models of the fluid-solid coupling,and a laboratory test of a rigid wedge sliding into water demonstrates that it can reasonably describe the dynamic behavior of a landslide-induced wave.A typical process of a block entering water and its energy variation are described and analyzed in detail.Further,the relationship between each characteristic parameter of the block and the waves is quantitatively investigated and fitted.The simulation results show that energy exchange between the block and the water is very rapid after the block collides with the water.The maximum kinetic energy,maximum velocity,duration,and side dimension of the waves mainly increase non-linearly with the above characteristic parameters of the block.The transformation ratio of the kinetic energy from the block to the water,the first wave,and the second wave are usually below 60%,45%,and 30%,respectively.The velocity of the block first decreases and then maintains a constant speed after entering the water.The displacement of the block increases linearly with the initial velocity and density of the block and exponentially increases with the block volume at different times.With the increase in the incident angle of the block,the kinetic energy and scale of the second wave increase correspondingly.

Design methods for antimicrobial peptides with improved performance

The recalcitrance of pathogens to traditional antibiotics has made treating and eradicating bacterial infections more difficult.In this regard,developing new antimicrobial agents to combat antibiotic-resistant strains has become a top priority.Antimicrobial peptides(AMPs),a ubiquitous class of naturally occurring compounds with broadspectrum antipathogenic activity,hold significant promise as an effective solution to the current antimicrobial resistance(AMR)crisis.Several AMPs have been identified and evaluated for their therapeutic application,with many already in the drug development pipeline.Their distinct properties,such as high target specificity,potency,and ability to bypass microbial resistance mechanisms,make AMPs a promising alternative to traditional antibiotics.Nonetheless,several challenges,such as high toxicity,lability to proteolytic degradation,low stability,poor pharmacokinetics,and high production costs,continue to hamper their clinical applicability.Therefore,recent research has focused on optimizing the properties of AMPs to improve their performance.By understanding the physicochemical properties of AMPs that correspond to their activity,such as amphipathicity,hydrophobicity,structural conformation,amino acid distribution,and composition,researchers can design AMPs with desired and improved performance.In this review,we highlight some of the key strategies used to optimize the performance of AMPs,including rational design and de novo synthesis.We also discuss the growing role of predictive computational tools,utilizing artificial intelligence and machine learning,in the design and synthesis of highly efficacious lead drug candidates.

Depolarization of Li-rich Mn-based oxide via electrochemically active Prussian blue interface providing superior rate capability

The high-rate cyclability of Li-rich Mn-based oxide(LMO)is highly limited by the electrochemical polarization resulting from the slow kinetic of the Li2MnO3 phase.Herein,the Prussian blue(PB)coating layer with specific redox potential is introduced as a functionalized interface to overcome the side effect and the escaping of O on the surface of LMO,especially its poor rate capability.In detail,the PB layer can restrict the large polarization of LMO by sharing overloaded current at a high rate due to the synchronous redox of PB and LMO.Consequently,an enhanced high rate performance with capacity retention of 87.8%over 300 cycles is obtained,which is superior to 50.5%of the pristine electrode.Such strategies on the high-rate cyclability of Li-rich Mn-based oxide compatible with good low-rate performances may attract great attention for pursuing durable performances.

Considering Regional Connectivity and Policy Factors in the Simulation of Land Use Change in New Areas:A Case Study of Nansha New District,China

Numerous emerging development areas worldwide are receiving attention;however,current research on land use change simulation primarily concentrates on cities,urban clusters,or larger scales.Moreover,there is a limited focus on understanding the impact of regional connectivity with surrounding cities and policy factors on land use change in these new areas.In this context,the present study utilizes a cellular automata(CA)model to investigate land use changes in the case of Nansha New District in Guangzhou,China.Three scenarios are examined,emphasizing conventional locational factors,policy considerations,and the influence of regional connectivity with surrounding cities.The results reveal several key findings:(1)Between 2015 and 2021,Nansha New District experienced significant land use changes,with the most notable shifts observed in cultivated land,water area,and construction land.(2)The comprehensive scenario exhibited the highest simulation accuracy,indicating that Nansha New District,as an emerging area,is notably influenced by policy factors and regional connectivity with surrounding cities.(3)Predictions for land use changes in Nansha by 2030,based on the scenario with the highest level of simulation accuracy,suggest an increase in the proportion of cultivated and forest land areas,alongside a decrease in the proportion of construction land and water area.This study contributes valuable insights to relevant studies and policymakers alike.